Aura is an open-source project designed to revolutionize blood bank management by moving from a reactive to a proactive, AI-driven decision-support system. It aims to reduce shortages, minimize wastage, and ultimately save lives.
Blood bank managers often operate in a reactive state. They know when inventory is currently low, but they lack the tools to accurately predict future shortages and take optimal, timely action. Current Blood Bank Management Systems (BBMS) are excellent systems of record, but not intelligent systems for proactive optimization.
Aura is not just another inventory tracker. It's a strategic assistant for blood bank managers. It integrates two key AI components to provide a complete, end-to-end solution.
Using time-series analysis and machine learning models (like SARIMA, Prophet, and LSTMs), Aura analyzes historical data, holidays, and other contextual factors to generate a reliable 7-10 day forecast of blood demand for each blood type.
This is Aura's core innovation. The forecast is fed into a Large Language Model (LLM) that acts as a reasoning engine. It assesses the predicted supply vs. demand and, instead of just raising an alert, it recommends specific, actionable, and optimized solutions, such as:
- "Launch a targeted SMS campaign to 150 previous O- donors."
- "Plan a donation drive at the local university campus this Thursday for maximum impact."
- "Draft a social media post highlighting the need for B+ blood."
The workflow is designed to be a seamless pipeline from raw data to actionable insight.
graph TD
A[Data Sources: Inventory Logs, Donor DB, Holidays] --> B{Data Pipeline};
B --> C[Predictive Engine: ML Forecasting];
C -- Forecast --> D[Prescriptive Engine: LLM Reasoning];
subgraph Core AI
C
D
end
D -- Recommendation --> E[Decision-Support Dashboard];
E -- Manager Approval --> F[Execute Action: SMS, Social Media, etc.];
Aura is built with a modern, open-source technology stack.
Follow these instructions to get a local copy up and running for development and testing purposes.
git clone [https://github.com/Vanshika-mahajan/aura-blood-inventory-ai.git](https://github.com/Vanshika-mahajan/aura-blood-inventory-ai.git) cd aura-blood-inventory-ai
python -m venv venv source venv/bin/activate # On Windows, use 'venv\Scripts\activate'
pip install -r requirements.txt
Create a .env file in the root directory by copying the example file.
cp .env.example .env
Fill in the necessary API keys or configuration details in your new .env file.
###5)Run the application:
Example for running a Streamlit dashboard
streamlit run app/dashboard.py
This project is divided into four key phases. We are currently in Phase 1.
[x] Identify core data requirements
[ ] Develop scripts to generate a realistic, synthetic dataset
[ ] Set up database schema
[ ] Implement baseline forecasting models (SARIMA, Prophet)
[ ] Experiment with ML models (XGBoost, LSTM)
[ ] Validate model accuracy and performance
[ ] Define actionable functions (e.g., send_sms, plan_drive)
[ ] Integrate an open-source LLM (e.g., Llama 3, Mistral)
[ ] Implement reasoning logic with function calling
[ ] Develop a backend API with FastAPI
[ ] Build an intuitive UI with Streamlit (for MVP)
[ ] Implement a feedback loop for recommendations
Contributions are what make the open-source community such an amazing place to learn, inspire, and create. Any contributions you make are greatly appreciated.
Please see the CONTRIBUTING.md file for details on our code of conduct and the process for submitting pull requests to us.
Distributed under the MIT License. See LICENSE for more information.
UCI Machine Learning Repository for the foundational dataset on donor behavior.
The open-source community for providing the tools that make this project possible.